Yelp & NYC Real Estate – Filtered Vs Unfiltered

If you’re a real estate agent in NYC (or any other business anywhere else apparently) you’ve heard the rumors about Yelp!: they’ll filter good reviews and let bad reviews go public about you or your company. Of course that’s unless you decide to advertise with them, and then they can make all those bad reviews go away. I’ve heard many agents/brokers complain about their reviews, both filtered and unfiltered. This is a look into what Yelp! does and doesn’t do and how businesses are affected by Yelp!’s rating practices.

Let’s start by stating that the business of reviews is complex. Yelp! has more than 67 million cumulative reviews, of which about 75% make it through to their “recommended” section on Yelp!s website.  This means that millions of other reviews are deemed “not recommended” for various reasons according to Yelp!’s algorithm. Yelp! states that it is their algorithm that makes all decisions regarding which reviews pass through filters, so we will start with examining how it works. Because Yelp! doesn’t release any information about their algorithm, we have to dig in to see how it works. The “recommendation software” as it is known uses dozens of points of information to see what constitutes a real review and what doesn’t, according to their content guidelines and terms of service.

Below are highlights from a survey of 50 NYC real estate brokerages, broken down by ratings: Recommended (unfiltered) and Not Recommended (filtered) for 2014.

There were a total of 1,576 reviews on the Yelp! website for the brokerages.
yelp_totalreviews   How the recommendations are distributed by rating/stars:   yelp_starreviews yelp_starreviews2   Total Ratings and Reviews (Recommended and Not Recommended)   yelp_totalreviews_pn   There were/are 8 Brokerages advertising on Yelp of the 50. Below is how they are rated:   yelp_advertiser_ratings  In looking at all the data, there doesn’t seem to be anything that indicates that Yelp! does indeed control reviews or ratings in the instance of influencing brokerages to advertise. While the 8 brokerages cited above does give a hint of outliers, there isn’t any evidence to suggest this goes on. You can also see this link and read more below. Example of a Yelp advertisement: SpaceMarketingShop_Yelp_Ad   Most & Least Recommended Brokerages: yelp_worstbroker_ratings What the Numbers Say

5 Star reviews are the most common, with reviewers leaving more than 1100+ ratings.  The runner-up is 1 star reviews which come in at 284 collectively. Out of the recommended reviews for both 5 star and 1 star reviews, 70% are positive with 30% being negative. Three star reviews are the least used by reviewers.

60% of brokerages had more positive ratings than negative.   yelp_rnr_   Reasons for Not Recommended Ratings (filtered reviews): There are many reasons a review gets “Not Recommended” aka filtered or flagged (see below).  To see what kind of control I had over my own reviews, I set up an account at Yelp!. I noticed a few things, mainly that providing misleading information or limited information will lead to a filtered review. I also noticed that I can alter and delete my reviews. So it’s possible that when asking clients to write a review for your services, they may or may not use their primary email address or add critical information for an account to post unfiltered reviews if they already don’t have a Yelp! account. Reviews Disappearing  I have experienced this first hand; a client had left me a positive review which was displayed on my Yelp! page. I didn’t solicit the review and when my client informed me about it, I thanked him for the review. After some time went by (6 months to 1 year) I went back to see if I had any additional reviews. Well, as it turned out, I didn’t have any reviews at all! My client’s review had disappeared from the recommended list and found its way to the not recommended list. I noticed that I was his only review. My client isn’t socially active online (Twitter, Facebook, Instagram, etc.) or a regular Yelp! user for that matter. This leads me to believe that Yelp!’s algorithm will, after some time, depending on the users profile, start filtering out older accounts with limited reviews. Understanding the Algorithm[table width =”100%” style =” table-hover” responsive =”true”] [table_head] [/table_head] [table_body] [table_row] [row_column]Activity (web/app)[/row_column] [row_column]Profile completion[/row_column] [row_column]Social connectivity[/row_column] [row_column]Social verification[/row_column] [/table_row] [table_row] [row_column]Amount of reviews[/row_column] [row_column]Amount of friends[/row_column] [row_column]Reviewed content (Details)[/row_column] [row_column]Review votes[/row_column] [/table_row] [table_row] [row_column]Flag profile[/row_column] [row_column]Review Ratio[/row_column] [row_column]Review feedback[/row_column] [row_column]Geolocation[/row_column] [/table_row] [table_row] [row_column]Content Guidelines[/row_column] [row_column]Terms of Service[/row_column] [row_column]Technical aspects[/row_column] [row_column]Other misc values[/row_column] [/table_row] [/table_body] [/table] Activity has to be one of the biggest contributing factors to reviews being filtered and unfiltered. This is, after all, what Yelp! Wants–active users. Without activity, Yelp! can’t sell advertising or business services. So, Yelp! rewards active users with badges, parties, and statuses. The above algorithm values is an educated guess on how Yelp!’s recommendation software works. How Yelp Filters Results Screen Shot 2014-12-29 at 11.16.08 AM

As you can see, the default “sorting” of reviews is Yelp!’s own method of displaying which reviews users see first. Since Yelp! doesn’t disclose how these reviews make their way onto the list and what constitutes the “Yelp! Sort,” it doesn’t make sense that this be the default view for users visiting business pages on the Yelp! website. It is safe to assume that these reviews have other reasons for being displayed.

While I think the date filter should be the most important factor when reading reviews, this isn’t the case. Reviews from the past can be a bit misleading. A complaint from, say, 3 years ago shouldn’t hang over a business’ head, especially if they have newer reviews which would more accurately reflect the current  consumer perception of customer service.

I do think that this is one way Yelp! can manipulate (for lack of a better word) reviews without actually changing anything. There is nothing wrong with this, after all the ninth circuit court said that they have the right to do what they want with reviews on the site. This is also reflected under their TOS. Oversight One of the biggest problems when it comes to reviews is that there isn’t any link to the company being reviewed or the reviewer. This gap–let’s call it the lost connection–is also a huge issue. How does a company know who the reviewer is? It could easily be a competing business, ex-employee, or simply the wrong business being reviewed.  It could even the business itself, paying for inflated reviews. Amazon and other direct service companies do have the advantage of selling products to consumers with profiles linked to reviews, making it easier to read honest reviews from actual buyers. This isn’t coming to Yelp! any time in the future just because of the pure investment and logistics that would allow such “verified” reviews. Takeaways for Real Estate Brokerages on Yelp

Brokerages should encourage agents to start their own Yelp! Page.  Doing so would reduce negative ratings on the company’s page. This is not to deflect negative ratings but rather to encourage agent responsibility and top performance.  Also, agents are constantly moving from brokerage to brokerage, so having their own pages would allow their reviews to move with them.

Brokerages should respond to negative reviews—actually, they should respond to pretty much all reviews. I noticed that many managers from brokerages would respond to reviews and some had positive effects in fostering and strengthening client relationships. Conclusion While there will always be some suspicion on how Yelp! manages their reviews, I don’t see anything that points to outright manipulation. Overall, the numbers suggest that more reviewers will leave a positive review than a negative one. [notification type=”alert-info” close=”false” ] Other Articles & Sources: ABC7 News  Harvard Study (Fake it till you make it) NYPost Ninth Circuit Court of Appeals Opinion Yelp recommendation software Notes: Yelp reviews (Recommended and Not Recommended) were compiled the first week January 2015.[/notification]

1 comment
Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts